Troubleshooting

The common snags and the fix for each. Still stuck? Compare notes in Discussions.

The bridge can't connect to the gateway

• Use the IP from a network scan, not the one shown in the mySigen app (see Getting started).
• Check the local data feed is still switched on in mySigen, and that the unit ID is 247 (the default).
• Use the Scan and Test connection buttons in Settings → Gateway to check from the bridge's side.
• Make sure the gateway didn't change address. Give it a fixed address on your router if you haven't.

Apple Home can't find the accessory, or is slow

• The container must run with host networking (the bundled compose.yaml already does this).
• On a NAS, or any host with extra Docker networks, set HOMEKIT_BIND to your network interface (e.g. eth0) or its IP, so the bridge doesn't advertise on unreachable internal 172.x addresses.
• Your phone must be on the same network as the host. Apple's local discovery doesn't cross network segments without help.

The Home app shows degrees instead of kilowatts

That's by design; the Home app has no power sensor, so the bridge borrows a temperature one. See Reading the numbers.

The dashboard tiles froze

• The footer shows the connection state and last update. If the gateway dropped, the bridge keeps the last known values and reconnects on its own.
• If a home-screen app on iOS has gone stale, tap the dashboard title to reload it.

No outdoor temperature in the header

Weather needs outbound internet (Open-Meteo). The first run works out where you are from the server's public address; set LATITUDE / LONGITUDE in settings if it guessed wrong, or turn weather off.

The cost figures look wrong

They're an estimate from the rates you enter, not a bill. Check Settings → Tariff has the right rates, windows (in local time), and currency, and that enough history has built up for today. See Tariffs & cost for what it can't model.

A settings change didn't apply

Almost everything applies live, but the Apple Home fields and the server port are marked "requires restart". A docker compose restart picks them up.

I changed .env but nothing happened

.env only seeds the first boot. Once you've saved settings in the dashboard, those win. Change it in Settings instead, or use Reset to fall back to the .env seeds.

Cloudflare Tunnel logs fill with "control stream encountered a failure"

cloudflared defaults to a transport (QUIC over UDP) that some hosts (a Synology NAS is the usual culprit) can't give enough buffer for. One of the four edge connections then wedges in a reconnect loop while the other three carry traffic, so the tunnel still serves but the logs fill with errors.

The fix is to force TCP instead. Add TUNNEL_TRANSPORT_PROTOCOL=http2 to the cloudflared service's environment, then recreate the container. A startup line about the UDP receive buffer being smaller than wanted is harmless on this setting.